← 返回 Skills 市场
385
总下载
0
收藏
1
当前安装
1
版本数
在 OpenClaw 中安装
/install companion-lobster
功能描述
陪伴型小龙虾 - 陪你刷抖音、看电影、听音乐、聊天分享的 AI 伙伴
使用说明 (SKILL.md)
🦞 陪伴型小龙虾 (Companion Lobster)
陪你刷抖音、看电影、听音乐、聊天分享的 AI 伙伴
✨ 核心功能
📱 刷抖音
- 发送抖音链接 → 获取视频内容
- 分析视频亮点,分享感受
- 推荐相关视频
- 一起评论区"冲浪"
🎬 看电影
- 输入电影名称或链接
- 讨论剧情、角色、主题
- 分享观后感
- 推荐相似电影
🎵 听音乐
- 分享歌曲链接或名称
- 讨论歌词、旋律、歌手
- 推荐相似风格音乐
- 分享音乐故事
💬 日常陪伴
- 随时聊天分享
- 记住你的喜好
- 主动推荐内容
- 情感陪伴支持
🚀 使用方法
安装
git clone https://github.com/adminlove520/companion-lobster.git
cd companion-lobster
npm install
使用场景
刷抖音
主人:https://v.douyin.com/xxx
小龙虾:分析视频内容 + 分享感受 + 推荐相关
看电影
主人:想看《星际穿越》
小龙虾:推荐电影 + 讨论剧情 + 分享观后感
听音乐
主人:这首歌真好听
小龙虾:推荐相似歌曲 + 讨论歌词 + 分享故事
🎯 交互示例
抖音分享
主人:https://v.douyin.com/mpJJ9N15jk8/
小龙虾:这个视频太有意思了!让我给你分析一下...
电影推荐
主人:有什么电影推荐吗?
小龙虾:看你喜欢什么类型~科幻?爱情?还是悬疑?
音乐分享
主人:听这首歌
小龙虾:这首太经典了!这句歌词我最喜欢...
🔧 技术实现
- 获取抖音/视频内容
- 电影/音乐信息检索
- AI 生成陪伴式回复
- 记住用户偏好
📝 更新日志
See CHANGELOG.md
📄 许可证
MIT
🦞 陪你度过每一个美好时刻!
安全使用建议
This package is not obviously malicious, but it is inconsistent: the docs promise networked features (Douyin/video fetching, movie/music lookups) while the shipped code is purely local stubs. Before installing or running npm install: 1) Inspect the repository (package.json) and any dependencies for postinstall scripts and network code; 2) Verify the upstream GitHub repo and its history/maintainer; 3) Run npm install in a sandboxed environment or container to limit risk; 4) If you expect live content fetching, confirm which APIs/endpoints and credentials (if any) the skill will use and whether those are safe to provide; 5) Consider reaching out to the author or choosing a skill with implemented, auditable network integration. If you only need local conversational stubs, the included code is harmless; do not assume the advertised external fetch capabilities are implemented.
功能分析
Type: OpenClaw Skill
Name: companion-lobster
Version: 1.0.0
The skill bundle provides a boilerplate template for a 'Companion Lobster' AI assistant designed to discuss Douyin videos, movies, and music. The core logic in `companion.js` consists entirely of placeholders and basic state management for user preferences, lacking any actual network requests, file system access, or sensitive data handling. No malicious instructions or prompt injection attempts were found in `SKILL.md` or the documentation.
能力评估
Purpose & Capability
The SKILL.md and README claim active features like "获取抖音/视频内容" and searching movie/music info. The included companion.js contains only local stub implementations that return placeholder strings and perform no network calls, API integration, or credential handling. That means the declared capabilities are not implemented in the shipped code and would require additional packages or external services to work — an incoherence between advertised purpose and actual shipped capability.
Instruction Scope
Runtime instructions in SKILL.md are limited (git clone, npm install, examples of usage) and do not instruct reading unrelated system files or exporting environment secrets. However the instructions are vague about how external content is fetched, giving the agent broad discretion to '获取抖音/视频内容' without specifying endpoints or credentials — that open-endedness can mask later addition of network calls or credentials in upstream code or dependencies.
Install Mechanism
There is no formal install spec in the skill bundle (instruction-only), but README/SKILL.md recommend 'git clone https://github.com/adminlove520/companion-lobster.git' and 'npm install'. The referenced GitHub repo/owner is unknown in this package metadata. Running npm install on unknown code can execute postinstall scripts and pull third-party packages — a moderate risk. The included bundle itself does not show any downloaded artifacts or installers.
Credentials
The skill declares no required environment variables, no credentials, and no config paths. The code does not access process.env or other secrets. Requested environment access is proportionate (none).
Persistence & Privilege
The skill is not forced-always (always:false) and does not request elevated persistent presence. It does not modify other skills or system settings in the provided files.
如何使用
- 确保已安装 OpenClaw(本地或 Docker 部署)
- 在对话框中输入安装命令:
/install companion-lobster - 安装完成后,直接呼叫该 Skill 的名称或使用
/companion-lobster触发 - 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
companion-lobster 1.0.0
- 首次发布:全新上线陪伴型小龙虾 AI 伙伴。
- 支持抖音视频分析、评论和推荐。
- 可陪伴看电影、讨论剧情与推荐相似影片。
- 支持音乐分享、讨论及推荐功能。
- 具备日常聊天、内容主动推荐和情感陪伴能力。
元数据
常见问题
companion-lobster 是什么?
陪伴型小龙虾 - 陪你刷抖音、看电影、听音乐、聊天分享的 AI 伙伴. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 385 次。
如何安装 companion-lobster?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install companion-lobster」即可一键安装,无需额外配置。
companion-lobster 是免费的吗?
是的,companion-lobster 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
companion-lobster 支持哪些平台?
companion-lobster 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。
谁开发了 companion-lobster?
由 Anonymous(@adminlove520)开发并维护,当前版本 v1.0.0。
推荐 Skills